# Getting Started¶

## Installation¶

alibi-detect can be installed from PyPI:

pip install alibi-detect


## Features¶

alibi-detect is a Python package focused on outlier, adversarial and concept drift detection. The package aims to cover both online and offline detectors for tabular data, text, images and time series. The outlier detection methods should allow the user to identify global, contextual and collective outliers.

To get a list of respectively the latest outlier and adversarial detection algorithms, you can type:

import alibi_detect
alibi_detect.od.__all__

['OutlierAEGMM',
'IForest',
'Mahalanobis',
'OutlierAE',
'OutlierVAE',
'OutlierVAEGMM',
'OutlierProphet',  # requires prophet: pip install alibi-detect[prophet]
'OutlierSeq2Seq',
'SpectralResidual',
'LLR']

alibi_detect.ad.__all__

['AdversarialAE',
'ModelDistillation']

alibi_detect.cd.__all__

['ChiSquareDrift',
'ClassifierDrift',
'KSDrift',
'MMDDrift',
'TabularDrift']


Summary tables highlighting the practical use cases for all the algorithms can be found here.

For detailed information on the outlier detectors:

And data drift:

## Basic Usage¶

We will use the VAE outlier detector to illustrate the usage of outlier and adversarial detectors in alibi-detect.

First, we import the detector:

from alibi_detect.od import OutlierVAE


Then we initialize it by passing it the necessary arguments:

od = OutlierVAE(
threshold=0.1,
encoder_net=encoder_net,
decoder_net=decoder_net,
latent_dim=1024
)


Some detectors require an additional .fit step using training data:

od.fit(X_train)


The detectors can be saved or loaded as follows:

from alibi_detect.utils.saving import save_detector, load_detector

filepath = './my_detector/'
save_detector(od, filepath)

preds = od.predict(X_test)

The predictions are returned in a dictionary with as keys meta and data. meta contains the detector’s metadata while data is in itself a dictionary with the actual predictions. It has either is_outlier, is_adversarial or is_drift (filled with 0’s and 1’s) as well as optional instance_score, feature_score or p_value as keys with numpy arrays as values.